I'm continuing work on disassembling and commenting the code on a a game cartridge. It's slow going, because I'm filling gaps in my modest 6502 assembly language programming as I go.

One of the oddball things I noticed while looking at the code is this this:



That looks like it's enabling writes to the cassette data line. But, I haven't spotted any writes to the cassette buffer. So, unless I've missed something there is no data actually being written to the cassette port itself.

Is this common in cartridge cold start routines? If so, why? Is it possible this is debug code intended to allow the cassette to be used as, well, something like a core dump target?

The 6510's "processor port" is not only used for the cassette interface. It also has some control lines that are used internally to the C64, particularly with respect to altering the memory map.

In particular, bits 0 and 1 (selected by a mask of #$03) control the /LORAM and /HIRAM lines.

Right, I follow that the "processor port" [pin] is used for lots of things. What I failed to account for was zero indexed bit-masking. As a consequence, I conflated $03 with setting bit-flag #3.

From Mapping the C64...



So, $03 (%00000011) would set /LORAM =1 (True), /HIRAM=1 (True) as you say, and does nothing else. By contrast, if I the cassette port were really enabled in actual fact, the value $04 (%00000100) would have been sent to $01. This is obviously not the case.

Thanks for helping me clear the cobwebs on this one.

It looks like you're off by a bit. The Cassette Data Output line would be 00001000 = $08
$04 would be the CHAREN signal.

Doesn't matter for this discussion, but maybe prevent confusion in the future!